Changing What You Like: Modifying Contour Properties Shifts Aesthetic Valuations of Scenes

Abstract

To what extent do aesthetic experiences arise from the human ability to perceive and extract meaning from visual features? Ordinary scenes, such as a beach sunset, can elicit a sense of beauty in most observers. Although it appears that aesthetic responses can be shared among humans, little is known about the cognitive mechanisms that underlie this phenomenon. We developed a contour model of aesthetics that assigns values to visual properties in scenes, allowing us to predict aesthetic responses in adults from around the world. Through a series of experiments, we manipulate contours to increase or decrease aesthetic value while preserving scene semantic identity. Contour manipulations directly shift subjective aesthetic judgments. This provides the first experimental evidence for a causal relationship between contour properties and aesthetic valuation. Our findings support the notion that visual regularities underlie the human capacity to derive pleasure from visual information.

The mid-level vision toolbox for computing structural properties of real-world images

“…For each contour property, the following fields are added to the vecLD struct: We have used these contour properties previously to explain behavior (Walther and Shen, 2014;Wilder et al, 2018) and neural mechanisms of scene categorization (Choo and Walther, 2016). We have also related statistics of these properties to the emotional content of scenes and generated artificial images with specific property combinations to elicit emotional responses (Damiano et al, 2021a), as well as to esthetic pleasure (Farzanfar and Walther, 2023).…”

“…We have used these functions to investigate the role of MAT-based features for computer vision (Rezanejad et al, 2019(Rezanejad et al, , 2023, eye movements (Damiano et al, 2019), neural representations of symmetry (Wilder et al, 2022), to investigate esthetic appeal (Damiano et al, 2021b;Farzanfar and Walther, 2023) and image memorability (Han et al, 2023).…”

“…splitLDbyStatsModel: splits the contours by the output of a statistical model, trained with the contour and MAT properties as its features. This function has been used to split contours according to predicted esthetic appeal (Farzanfar and Walther, 2023) or according to predicted memorability (Han et al, 2023).…”